Reading one human genome once cost about a hundred million dollars. It now costs a few hundred.
We can read the code of life almost for free. We still cannot say what most of it does.
Nine in ten medicines that reach human trials fail. Biology is not yet an engineering discipline. It can become one.
In 2001 the first human genome cost around 95 million dollars to sequence. By 2022 the figure was near 500 dollars. No other measurement has fallen in price so fast: about two hundred thousand times in twenty years, far outpacing computer chips.
The same period added tools for writing as well as reading. Gene editing lets a researcher change a chosen letter in a living cell. Vaccines can be designed on a computer from a published sequence in days. The first therapies that edit a patient's own cells are approved and in use.
From the price of a jet airliner to the price of a pair of shoes.
Every cell in your body carries the same three billion letters and reads a different part of them. The path from letter to protein is understood in detail. The step that defeats us is the last one: how twenty thousand kinds of protein, regulating each other, add up to a living cell.
Source: International Human Genome Sequencing Consortium (2004); Nurk et al., Science (2022); Bianconi et al. (2013) for cell count · cited from memory, to verify
A child born in 1950 could expect to live about 46 years, averaged across the world. A child born today can expect about 73. Most of that gain came from clean water, vaccines, antibiotics and food, not from advanced medicine.
Those gains are far from evenly shared, and they are slowing where they were greatest. The diseases that remain, such as cancer, dementia and heart disease, are diseases of the body's own machinery going wrong over decades. They will not yield to simple things.
The dip is the pandemic. The long rise is the largest improvement in the human condition on record.
Source: United Nations, World Population Prospects 2024 · cited from memory, to verify
Medicine works across this whole range at once. A drug is a molecule a nanometre across, aimed at a protein five times its size, inside a cell ten thousand times larger, in a body a hundred thousand times larger again. Each step up is a different science.
Source: Milo and Phillips, Cell Biology by the Numbers (2015) · cited from memory, to verify
Of every hundred drug candidates that enter human trials, about eight are approved. The usual reason for failure is not safety. It is that the drug does not work, because the idea about the disease was wrong.
That is the mark of a field without a predictive theory. An aircraft engineer can calculate whether a wing will hold before building it. A biologist cannot yet calculate what a cell will do when one protein is changed, because a cell is thousands of parts regulating each other at once.
The cost follows from the failure rate. Developing one medicine takes ten to fifteen years, and every success pays for the failures behind it.
Most fail in the second phase, the first real test of whether the drug works.
Source: BIO, Informa and QLS, Clinical Development Success Rates and Contributing Factors 2011 to 2020 (2021) · cited from memory, to verify
Each box is as wide as the time it takes. The largest loss comes in the second phase of trials, years in, when a drug that is safe turns out not to work. A way to learn that at the start, in a model of the disease, would change the economics of medicine more than any single cure.
Source: BIO, Informa and QLS, Clinical Development Success Rates 2011 to 2020; phase durations are typical published ranges · cited from memory, to verify
Our aim is predictive biology: models of cells and tissues good enough that an experiment can be run on a computer first, and the laboratory used to check the answer. Cheap measurement, automated laboratories and learned models now make that a serious programme, not a slogan.
From there the direction is prevention. Detect disease years before symptoms, when it is cheapest to stop. Treat ageing itself as a process with causes that can be studied. And insist that what is learned reaches people everywhere, since the thirty years gained so far came from things that were cheap enough to share.
Three genes sit in a ring. Each one's protein switches the next gene off. This circuit was designed on paper and then built inside living bacteria in 2000. Change how strong the genes are, or silence one.
This is a statement of intent, not a record of achievement. Each step carries a label for how firm it is. Nothing marked proposed or speculative has been done yet.
A platform in the style of Anu and Dhi: the central machinery of the cell as working simulations, with every claim labelled.
A shared, testable simulation of a minimal bacterium that predicts the outcome of real knock-out experiments.
Design tools that cut the failure rate of new therapies by predicting which ideas about a disease are wrong.
Cheap, repeated measurement that catches cancer and dementia early enough to stop them, offered widely.
Healthy years extended by acting on the causes of ageing, and available in every country.
Every claim on this page is open to challenge in the Agora, in the Life & Medicine room.
Argue it in the Agora ↗